In July we began a polling project to measure public confidence in government institutions on national security matters on an ongoing basis. In this third monthly installment of that project we continue to monitor trends in the four core questions about confidence in government institutions, along with perceptions of the two political parties on national security that we began asking about in August. This month’s data suggests public confidence in government institutions when it comes to national security is trending downward, though only marginally. The data also reveal that the public has equally low confidence in both political Parties when it comes to national security, but with wide variations across gender and age groups.
Confidence in Institutions Dipped Down
In late September, we once again used Google Surveys to ask the four questions on confidence in government institutions that we asked in the previous two months, namely:
In September, the public’s confidence in all four institutions declined from the levels we saw in August. The courts saw the biggest decline in confidence, with their average score dropping by .14 on a scale of 1 (no confidence) to 5 (high confidence). Average confidence in Congress and the intelligence community dropped by .08, while confidence in the President dropped only marginally. For all four institutions, this reversed gains in confidence made in August. Both months’ changes may reflect natural variability in a generally stable level of confidence. It’s also possible that September’s results reflect a dispirited public reaction to repeated hurricanes.
Aside from the general downward trend in confidence, results from September largely mirrored the results from earlier months. The intelligence community remains the only institution with an average score above 3, indicating more Americans trust it than not. On the other hand, the public continues to have the least confidence in Congress. There also continues to be a wide gender divide on confidence in institutions. For example, men are twice as likely as women (34 percent to 17 percent, respectively) to report having high confidence in the President. Additionally, age continues to be an important factor in the public’s level of confidence in institutions. This is particularly true for the intelligence community, where those under 45 are nearly twice as likely as those 45 and older (18 percent to 11 percent, respectively) to report having no confidence in the intelligence community to protect U.S. national security.
Parties at Parity on National Security
In September, as in August, we also asked respondents about their perceptions of the political parties when it comes to national security, specifically:
In September the two parties had identical confidence levels of 2.7, unlike in August, when average confidence in the Republican Party (2.8) was marginally higher than the average confidence level for the Democratic Party (2.6). This resulted from public perceptions of Democrats improving slightly, and perceptions of Republicans dropping slightly. For both parties, the modal response was, by far, “no confidence.” Yet despite this low level of confidence, the public continues to have more confidence in the parties than in Congress, which had an average confidence level of just 2.51.
Younger and Female Respondents Trust Democrats More; Older and Male Respondents Trust Republicans More
As we found in August, there are marked differences in levels of party trust by age and gender. The average level of trust of the Democratic Party for those under 45 is 2.8, compared to just 2.4 for the Republican Party. Conversely, those 45 and older have a nearly inverse perception of the parties, with the Republican Party having an average confidence score of 2.9 and the Democratic Party scoring 2.5 with this age group. The differences between these age groups are particularly striking when just looking at the percentage of respondents saying they have “no confidence” in each party. For example, 41 percent of respondents 45 and older reported having “no confidence” in the Democratic Party, while a much smaller—though still sizeable—percentage (32 percent) of this age cohort reported “no confidence” in the Republican Party.
There is a similarly clearly defined difference between men and women when it comes to levels of confidence in the two parties on national security, with women being much more likely to support the Democratic Party than men (average confidence scores were 2.7 and 2.6, respectively). The gender breakdown on confidence in the Republican Party to protect U.S. national security was even starker, with women having an average confidence score of just 2.5, and men’s average confidence at 2.9. Again, these gender differences are most evident in the modal category for both parties—“no confidence.” For example, while a third of men reported having no confidence in the Republican Party, more than four in ten women did so.
While these age and gender differences have remained relatively constant since we first asked this question in late August, the big takeaway from this month’s polling is that the public now sees little difference between the two parties when it comes to confidence in their ability to protect U.S. national security. Whether this is a product of the President’s seemingly chaotic approach to national security or whether the apparent Republican edge was always a matter of statistical noise is a question for debate, one that will become clearer as we continue to run these questions on a monthly basis and develop a sense of the regular ebb and flow of confidence in government institutions and political actors.
An additional note on this project: The folks at Google Surveys have decided to support this research by donating significant access to the platform for our use. Over the coming few months, therefore, we will be significantly ramping up the number of questions we are polling and expanding the diversity of questions asked. We are grateful to Google for allowing us to do this. By mutual agreement, Google will play no role in either decisions about polling questions or our editorial decisions as to how to characterize the results.